Year: 2021

Instagram head Adam Mosseri confirmed the company is testing a new feature called “Limits,” which would give users the ability to temporary lock down their accounts when they’re being targeted by a flood of harassment. The announcement of the new feature was made today during a video where Mosseri condemned the recent racism that took place on Instagram’s platform following the Euro 2020 final, and noted the company was working on improvements to both internal and customer-facing tools to help address this problem.

The company had previously commented on and condemned the racist abuse, which had seen England footballers Bukayo Saka, Marcus Rashford and Jadon Sancho viciously harassed by angry fans making racist comments after the team’s defeat earlier this month. Mosseri explained at the time the company was using technology to try to prioritize user reports, and it mistakenly marked some reports as benign comments instead of referring them to human moderators. One of the possible complications was that many of the harassing comments were using emoji, which Instagram’s systems may have struggled to understand given emoji can have different meanings in different contexts.

Today, Mosseri again acknowledged Instagram’s mistake and noted it has since fixed the issue. He said Instagram had been proactively sweeping the footballers’ comments, but hadn’t anticipated the wave of user reports.

He also pointed out that Instagram receives millions of user reports per day and even getting 1% of them wrong leads to tens of thousands of problematic posts that remain on the platform in error.

Mosseri then mentioned several user-facing tools that could help people deal with harassment more directly on their own accounts to prevent abuse. This includes Instagram’s tools like Block and Restrict. The latter is tool that allows users to approve a user’s comments before anyone else sees them or read someone’s messages without sending read receipts. Another more recently added tool called Hidden Words lets users block certain keywords in both comments and direct messages.

He added that Instagram was also testing a new tool called Limits, which would allow users to lock down their accounts in a “moment of risk.” This feature could have helped the footballers, as it would have offered simple settings to limit unwanted comments and reactions.

The feature had been spotted earlier this month by social media consultant Matt Navarra, who shared screenshots of how it worked, but Instagram had yet to formally announce it.

In the images that were shared, users with the feature would find a new section called Limits in Instagram’s privacy controls which explained that they could temporarily limits comments and messages from specific groups of followers.

Users could then toggle on or off groups to limit, including recent followers and accounts that are not following you, as these could include accounts that were spam or those created just to harass you. As is often the case, when there’s a flood of incoming abuse, it will not come from an account’s longtime followers, but rather from newcomers who have sought out the account just to harass them.

The feature will also allow users to set a duration for the Limits in terms of a number of days or even weeks.

An Instagram spokesperson also confirmed the feature worked as the images show, noting it would be a tool that would help people manage “intense instances of harassment or abuse.”

“Maybe you’re in high school and you are going through break up or you just switched schools. Or maybe you are professional footballer and you’re receiving a lot of harassment,” explained Mosseri, when detailing how Limits could be useful in different situations. “Whatever it is, we know that people sometimes are in temporary moments of real risk of pain, and we want to give them tools to protect themselves in those situations,” he added.

Instagram declined to say when the feature would become publicly available, but noted it’s being tested on mobile in select countries for the time being.

Over the years, I’ve had a front-row seat to the future of technology.

In my role at Y Combinator as director of admissions, I saw hundreds of startup pitches. Many shared a particular attribute: They followed the path of quickly growing users and monetizing the data extracted from the user.

As time went on, I began to see the full picture of what our technologies were creating: A “Minority Report” world where our every move is tracked and monetized. Some companies, like Facebook, lived by the mantra “move fast, break things.” Not only did they break things, they failed us by propagating disinformation and propaganda that, ultimately, cost some people their lives.

And that happened because of a growth-at-all-costs mindset. Some of the biggest consumer-facing Silicon Valley companies in the 21st century flourished by using data to sell ads with little or no consideration for user privacy or security. We have some of the brightest minds in technology; if we really wanted to, we could change things so that, at the very least, people wouldn’t have to worry about privacy and the security of their information.

We could move toward a model where people have more control over their own data and where Silicon Valley explores innovations in privacy and data security. While there are multiple long-term approaches and potential new business models to explore, there are ways to approach a privacy-first mindset in the near term. Here are a couple of ways to start moving toward a future in which people can have control over their data.

Workplace applications should lead the charge in enabling more secure identity technologies

We need to approach technology by consciously designing a future where technology works for humans, businesses and society in a secure and ethical way.

Approaching technological growth without understanding or considering the consequences has eroded trust in Silicon Valley. We must do better — and we can start in the workplace by better protecting personal data through self-sovereign identity, an approach that gives people control and ownership over their digital identity.

Using the workplace as a starting point for better privacy and security of people’s digital identities makes sense because many technologies that have been widely adopted — think personal computers, the internet, mobile phones and email — started out in the workplace before they became household technologies, thereby inheriting the foundational principles. With a return to office life on the horizon, there’s no better time than now to reexamine how we might adopt new practices in our workplaces.

So how would employers do this? For starters, they can use the return to office as an impetus for contactless access and digital IDs, which protect against physical and digital data breaches, the latter of which are becoming more common.

Employees could enter offices through their digital IDs, or tokenized IDs, which are stored securely on their phones. They will no longer need to use plastic cards with their personal information and photo imprinted on them, which are easy to fake or duplicate, improving security for both the employer and employee.

Contactless access isn’t a big leap nowadays, either. The pandemic primed us for digital identification — because the use of contactless payment accelerated due to COVID, the change to contactless ID will be seamless for many.

Invest in critical privacy-centric infrastructure

Tokenized identification puts the power in the user’s hands. This is crucial not just for workplace access and identity, but for a host of other, even more important reasons. Tokenized digital IDs are encrypted and can only be used once, making it nearly impossible for anyone to view the data included in the digital ID should the system be breached. It’s like Signal, but for your digital IDs.

As even more sophisticated technologies roll out, more personal data will be produced (and that means more data is vulnerable). It’s not just our driver’s licenses, credit cards or Social Security numbers we must worry about. Our biometrics and personal health-related data, like our medical records, are increasingly online and accessed for verification purposes. Encrypted digital IDs are incredibly important because of the prevalence of hacking and identity theft. Without tokenized digital IDs, we are all vulnerable.

We saw what happened with the Colonial Pipeline ransomware attack recently. It crippled a large portion of the U.S. pipeline system for weeks, showing that critical parts of our infrastructure are extremely vulnerable to breaches.

Ultimately, we need to think about making technology that serves humanity, not vice versa. We also need to ask ourselves if the technology we create is beneficial not just to the user, but to society in general. One way to build technology that better serves humanity is to ensure that it protects users and their values. Self-sovereign identity will be key in our future as other technologies arise. Among other things, we will see our digital wallets house far more than just credit cards, making the need for secure digital IDs more critical. Most importantly, people and companies just need control over their own data, period.

Given the broader general awareness of privacy and security in recent years, employers must take the threat of personal-data vulnerability seriously and lead the way in self-sovereign identity. Through the initial step of contactless access and digital IDs in the workplace, we can begin to inch closer toward a more secure future, at least in terms of our own data and identity.

Playdate, the adorable whimsy-and-nostalgia-box/handheld game system built by Panic (with some help from Teenage Engineering), has taken one more big step toward reality: it has an official pre-order date. And it’s soon!

The company announced this morning that pre-orders for the handheld will go live on July 29th at 10 a.m pacific.

Looking to get one from the first batch? Here’s the other stuff you need to know:

• The handheld will cost $179, and they’ll be selling an optional case accessory for $29. They’re offering both as a bundle for $199, saving you a couple bucks if you already know you want both. No word yet on when the previously announced docking station will go on sale.
• It sounds like the actual ship date still isn’t fully locked in, but Panic says the first batch (20,000 units or so) should start going out “towards the end of the year,” with additional units going out in 2022. The company stresses that they’re not capping pre-orders so they can’t really “sell out”, but ordering earlier means getting it sooner.
• Pre-orders will be capped at two per person.

Panic first announced the Playdate in 2019. Games on the Playdate are released in “seasons”; in season one, two new titles will be released each week for twelve weeks. As experimental as it is charming, Panic is pretty open about what to expect of the titles. From their product page: “Some are short. Some are long. Will you love them all? Probably not. Will you have a great time trying them? Absolutely.”

Rivian, the Amazon-backed electric automaker that aims to be the first to bring an EV pickup truck to market, plans to open a second U.S. manufacturing factory, sources told TechCrunch, confirming an earlier report from Reuters.

Rivian wouldn’t elaborate on when it planned to build the factory, but did confirm it was in the process of identifying a site for a second plant. Reuters reported that the factory, dubbed Project Tera, would also include battery cell production, a detail that would drive up the cost and size of the factory.

“While it’s early in an evolving process, Rivian is exploring locations for a second U.S. manufacturing facility,” spokesperson Amy Mast said in an emailed statement. “We look forward to working with a supportive, technology-forward community in order to create a partnership as strong as the one we have with Normal, Illinois.”

The news comes a week after Rivian CEO RJ Scaringe sent a letter to customers that the company was pushing back deliveries of its long-awaited R1T electric pickup truck and R1S SUV several more months due to delays in production caused by “cascading impacts of the pandemic,” particularly the ongoing global shortage of semiconductor chips. The R1T deliveries will begin in September with the R1S to follow “shortly,” Scaringe wrote in the message.

Rivian plans to assemble its consumer products — the R1T, R1S — as well as the commercial delivery vans slated for Amazon at its factory in Normal, Illinois. That factory, which once produced the Mitsubishi Eclipse through a joint venture between Mitsubishi and Chrysler Corporation, has been completely updated and expanded.

The Normal factory has two separate production lines producing vehicles. One is dedicated for the R1 vehicles and other line is for its commercial vans. Amazon ordered 100,000 of these vans, with deliveries starting in 2021.

The automaker has raised more than $8 billion from a diverse set of backers that includes Ford, Cox Automotive, T. Rowe Price Associates Inc., Fidelity Management and Research Company, Amazon’s Climate Pledge Fund, Coatue and D1 Capital Partners. That capital will be needed to keep its 7,000-person workforce running and while building an assembly factory, a project that will cost at least $1 billion if it follows industry estimates.

With more venture funding flowing into the startup ecosystem than ever before, there’s never been a better time to be a growth expert.

At TechCrunch Early Stage: Marketing and Fundraising earlier this month, Greylock Partners’ Mike Duboe dug into a number of lessons and pieces of wisdom he’s picked up leading growth at a number of high-growth startups, including StitchFix. His advice spanned hiring, structure and analysis, with plenty of recommendations for where growth teams should be focusing their attention and resources.

How to define growth

Before Duboe’s presentation kicked off, he spent some time zeroing in on a definition of growth, which he cautioned can mean many different things at many different companies. Being so context-dependent means that “being good at growth” is more dependent on honing capabilities rather than following a list of best practices.

Growth is something that’s blatantly obvious and poorly defined in the startup world, so I do think it’s important to give a preamble to all of this stuff. First and foremost, growth is very context dependent; some teams treat it as a product function, others marketing, some sales or “other.” Some companies will do growth with a dedicated growth team; others have abandoned the team but still do it equally well. Some companies will goal growth teams purely on acquisition, others will deploy them against retention or other metrics. So, taking a step back from that, I define growth as a function that accelerates a company’s pace of learning.

Growth is everyone’s job; if a bunch of people in the company are working on one problem, and it’s just someone off in the corner working on growth, you probably failed at setting up the org correctly.  (Timestamp: 1:11)

While growth is good, growing something that is unsustainable is an intense waste of time and money. Head of growth is often an early role that founders aim to fill, but Duboe cautioned early-stage entrepreneurs from focusing too heavily on growth before nailing the fundamentals.

I’ve seen many companies make the mistake of working on growth prior to nailing product-market fit. I think this mistake becomes even more common in an environment where there’s rampant VC funding, so while some of the discipline here is useful early on, I’d really encourage founders to be laser-focused on finding that fit before iterating on growth. (Timestamp: 2:29)

Where to focus growth energy

The bulk of Duboe’s presentation focused on laying out 10 of the “most poignant and generalizable” lessons in growth that he’s learned over the years, with lessons on focus, optimization and reflection.

Lesson 1: Distill your growth model (“business equation”)

Growth modeling and metric design — I view as the most fundamental part of growth. This does not require a growth team so any good head of growth should require some basic growth model to prioritize what to work on. (Timestamp: 3:09)

The first point Duboe touched on was one on how to visualize your growth opportunities using models, using an example from his past role leading growth at Tilt, where his team used user state models to determine where to direct resources and look for growth opportunities.

Lesson 2: Retention before acquisition

The second lesson is to prioritize retention before driving acquisition, a very obvious or intuitive lesson, but it’s also easy to forget given it’s typically less straightforward to figure out how to retain users versus acquiring new ones. (Timestamp: 4:19)

Retention is typically cheaper than acquiring wholly new users, Duboe noted, also highlighting how a startup focusing on retention can help them understand more about who their power users are and who exactly they should be building for.

Lesson 3: Embrace ideas from all corners, but triage

Bringing on new ideas is obviously a positive, but often ideas need guidelines to be helpful, and setting the right templates early on can help team members filter down their ideas while ensuring they meet the need of the organization.

Tesla will secure nickel from the commodity production giant BHP, the automaker’s latest move to secure direct sources of raw materials that are projected to surge in demand before the decade is out.

BHP’s Nickel West division will supply an undisclosed amount of the mineral from its mines in Western Australia. The two companies also agreed to work together to increase battery supply chain sustainability and to identify ways to decrease carbon emissions from their respective operations using energy storage paired with renewable energy.

Nickel is a key mineral in lithium-ion batteries, and a cornerstone of Tesla’s next-gen battery chemistry. While many lithium-ion batteries have cathodes made from nickel, manganese and cobalt, Tesla is taking a different tack. At Tesla’s Battery Day 2020, Musk said the automaker would invest in a nickel-rich, cobalt-free cathode for some models, citing greater energy density.

Tesla also hasn’t been shy about its own intention to increase battery cell production in the coming decade, aiming to produce 100 gigawatt hours of batteries by 2022, to a staggering 3 terawatt hours per year by 2030.

To that end, the company is moving fast to secure purchase agreements with leading nickel producers. Earlier this year, the automaker announced a partnership with a nickel producer in the French Pacific territory New Caledonia. Just a few months later, Tesla chairperson Robyn Denhlm confirmed that the company was looking to purchase around $1 billion per year in battery minerals from Australia alone.

Musk has repeatedly urged miners to produce more nickel. On an investment call last July, he told producers, “Tesla will give you a giant contract for a long period of time if you mine nickel efficiently and in an environmentally sensitive way.” At Battery Day, he reiterated his position: “In order to scale, we really need to make sure that we’re not constrained by total nickel availability,” he said. “I actually spoke with the CEOs of the biggest mining company in the world and said, ‘Please make more nickel, it’s very important.’”

But finding an environmentally friendly nickel source is a challenge. Some of that has to do with issues endemic to present-day recovery and smelting techniques; others are more directly manageable by mining companies. For example, nickel mining operations in Indonesia, the world’s largest producer of the metal, have come under fire for their reliance on coal and their waste disposal techniques.

BHP claims its operation is one of the most sustainable in the world, and Tesla’s decision to partner with it could be seen as something of a confirmation of that fact. The commodity producer in February said up to 50% of the electricity to power one of its nickel refineries would come from solar energy resources.

The vast majority of the world’s nickel supply is currently consumed by the steel industry. While nickel demand in the EV and energy storage sectors is currently relatively small, the International Energy Agency estimates that will increase more than 4,000% over the next 20 years – from 81 metric tons in 2020 to 3,352 metric tons by 2040.

Nickel West has historically been a tiny fraction of BHP’s overall business, dwarfed by its iron ore, copper and petroleum businesses. The commodity producer tried to sell Nickel West a number of times since around 2015, but it appears to have changed its tune with the forecasted groundswell of demand from the EV and energy storage sectors.

Industry analysts Benchmark Minerals estimated the deal with Tesla could be worth up to 18,000 tons of nickel annually.

A large chunk of the internet dropped offline on Thursday. Some of the most popular sites, apps and services were down, including UPS and FedEx (which have since come back online), Airbnb, Fidelity, and others are reporting Steam, LastPass, and the PlayStation Network are all experiencing downtime.

Many other websites around the world are also affected, including media outlets in Europe.

What appears to be the cause is an outage at Akamai, an internet security giant that provides networking and content delivery services to companies. At around 11am ET, Akamai reported an issue with its Edge DNS, a service that’s designed to keep websites, apps and services running smoothly and securely.

DNS services are critically important to how the internet works, so when things go wrong or there’s an outage, it has a knock-on effect to all of the customer websites and services that use it.

Akamai said it was “actively investigating the issue,” but when reached a spokesperson would not say if its outage was the cause of the disruption to other sites and services that are currently offline.

It’s not known what caused the incident but Akamai said it was already in recovery.

“We have implemented a fix for this issue, and based on current observations, the service is resuming normal operations. We will continue to monitor to ensure that the impact has been fully mitigated,” Akamai told TechCrunch.

It’s not the first time we’ve seen an outage this big. Last year Cloudflare, which also provides networking services to companies around the world, had a similar outage following a bug that caused major sites to stop loading, including Shopify, Discord and Politico. In November, Amazon’s cloud service also stumbled, which prevented it updating its own status page during the downtime. Online workspace startup Notion also had a high-profile outage this year, forcing the company to turn to Twitter to ask for help.

India’s central bank is considering launching a digital currency, according to a top executive, giving a clear indication of its intentions for the first time after previously stating that it was studying the idea.

T Rabi Sankar, the deputy governor of Reserve Bank of India, said at a conference today that the central bank is considering introducing the nation’s digital currency in a “phased” manner while legal changes are made to the South Asian nation’s foreign-exchange rules and IT laws.

The digital currency, which will be backed by sovereign, will lower the economy’s reliance on cash, enable cheaper and smoother international settlements, and protect people from the volatility of privacy cryptocurrencies, he said.

“Every idea has to wait for its time, and the time for CBDC [central bank digital currency] is near. We have carefully evaluated the risks,” he told an audience at a conference held by think-tank Vidhi Centre for Legal Policy.

Sankar said the central bank’s “endeavor is that as we move forward [with the plan],” so that India’s digital currency “can reiterate its leadership position in payment systems of the world.”

The top executive’s remarks follows European Central Bank saying last week that it will begin a 24-month “investigation phase” that, if successful, could lead to the creation of a digital euro by 2025.

Also last week, China’s central bank said its digital yuan trial had reached $5.3 billion in transaction value by the end of June.

“Central banks have increased their attention on digital currencies,” said Sankar. “CBDC will be in the arsenal of most if not all central banks in the world. A calibrated and nuanced approach will be considered at the drawing board as well as with stakeholder consultations,” he said, adding that the central bank has been exploring the benefits and risks of issuing a sovereign CBDC for “quite some time.”

“We have studied specific-purpose CBDCs proposed by different central banks around the world for wholesale and retail segments. The launch of a general-purpose CBDC for population scale is being considered, and RBI is working towards a phased introduction strategy and examining use cases with little or no disruption of India’s banking and monetary systems,” he said. “However, conducting pilots in wholesale and retail segments may be a possibility in near future.”

In his remarks, Sankar also hinted that the central bank hasn’t changed its stand on private cryptocurrencies such as bitcoin.

In 2018, an Indian government panel recommended banning all private cryptocurrencies and proposed up to 10 years of jail time for offenders. The panel also suggested the government to explore a digital version of the fiat currency and ways to implement it.

At the time, RBI said the move was necessary to curb “ring-fencing” of the country’s financial system. It had also argued that bitcoin and other cryptocurrencies cannot be treated as currencies as they are not made of metal or exist in physical form, nor were they stamped by the government.

“They are not commodities or claims on commodities as they have no intrinsic value; some claims that they are akin to gold clearly seem opportunistic,” Sankar said today.

The 2018 notice from the central bank sent a panic to several local startups and companies offering services to trade in cryptocurrency. Nearly all of them have either since closed shop, or pivoted to serve other markets.

This proposal was challenged by several exchanges and traders, who filed a lawsuit in the Supreme Court. The nation’s apex court ruled in their favor last year. This ruling was seen as “historic” but it has yet to impact the earlier circular on the policy level. In the meantime, the country has hinted that it plans to introduce a law to ban private cryptocurrencies.

In the agenda published on the lower house website earlier this year, a legislation sought to “prohibit all private cryptocurrencies in India,” but allow “for certain exceptions to promote the underlying technology [blockchain] of cryptocurrency and its uses.”

Gopuff, the “instant” grocery delivery startup that has been on an acquisition and expansion tear in the last several months to scale its business, is also racing to raise money to fuel those efforts. Documents uncovered by Prime Unicorn Index and shared with TechCrunch show that the startup has filed papers in Delaware to raise up to $750 million in a Series H round of funding — at a valuation of $13.5 billion if all shares are issued. While the company is not commenting on the filing, a well-placed source tells us that it’s actually closing as a $1 billion raise at a $15 billion valuation.

As with all Delaware filings, they only tell part of the story, so the company might ultimately raise more or less before the round closes. (And in this case it looks like “more.”)

For some funding context, it was only in March that Gopuff raised $1.15 billion at an $8.9 billion valuation. And that round came just months after a $380 million round (at a $3.8 billion) valuation. With Gopuff’s instant grocery model comes instant funding, it seems: together the three rounds would total around $2.5 billion in funding in the space of 10 months. (Investors in the company’s previous rounds have included Accel, D1 Capital Partners, Fidelity Management and Research Company, Baillie Gifford, Eldridge, Reinvent Capital, Luxor Capital, and SoftBank.)

Much like the investment race in the transportation-on-demand market, a large part of the fundraising in instant grocery seems to be aimed at scaling as fast as possible to build technological, operational and customer moats.

So for Gopuff, some of the money it’s raised so far has been used to expand organically. That is, it’s investing to acquire new customers and build out its infrastructure — riders, “dark” stores stocked with their products, and most recently “Gopuff kitchens” — within the 650+ cities in the U.S. where it already operates its $1.95 flat fee “in minutes” delivery service. It will likely be doing so at a particularly fast pace, considering that others like DoorDash are also moving in to compete in earnest around the same model for quick deliveries of a limited selection of food and drinks, home essentials, and over-the-counter medication.

But alongside that, some of the cash it is amassing is also being used for acquisitions. So far, these have been limited to the U.S. and to expand Gopuff’s breadth in that market. It bought alcohol retailer BevMo for $350 million in November 2020; and in June of this year Gopuff acquired logistics tech company rideOS for $115 million.

The next stage of that acquisition process looks like it may be focused on snapping up similar companies in key markets where Gopuff wants to be in the future, particularly internationally, as it works to fill out a reported ambition of reaching $1 billion in revenues this year (3x last year’s numbers).

In June, there were rumors around that Gopuff had approached Flink, an instant grocery player in Germany. While that has not gone anywhere (yet?), well-placed sources have told us — and, it seems, others — that Gopuff is also casting its international eye on England, engaging in discussions to acquire two different instant delivery companies based out of London, first Fancy back in February, and more recently, Dija.

Gopuff also declined to coment on Dija but we have multiple, well-placed sources telling us it’s in the works.

London is a hugely competitive market for instant grocery delivery at the moment — not least because it is dense and often hard to get around, has demonstrated a strong consumer appetite for on-demand delivery services, and has a population of younger people with a decent amount of disposable income to pay a little more for convenience.

That speaks of opportunity, but also possibly too many hopefuls as well. In addition to Dija and Fance, we have Turkey’s Getir, backed by Sequoia and a number of others on an ambitious international roll at the moment; Gorillas (like Flink, from Berlin); Zap; and Weezy — all offering “instant” grocery delivery. And these are just the standalone, newer startups. Still to come: established restaurant delivery players like Deliveroo that might also throw their hats into the ring.

Perhaps unsurprisingly, given that field, we’ve heard that Dija has been struggling to raise more money, and that led to the startup looking for buyers as an alternative.

That is a trend that’s playing out elsewhere too: In Spain Getir earlier this month acquired Blok, another new instant player that was struggling to get investors on board. We confirmed with well-placed sources that Dija had also talked with Getir in this context (that didn’t go anywhere) before Gopuff entered the picture. There will likely be more of these.

“It’s going to be a bloodbath,” is how one big investor recently described the instant grocery market to me.

Given that online grocery remains a relatively minor part of the market — even with the pandemic and its habit-changing impact on e-commerce, it’s still under 10% of sales, even in the most adoption-friendly cities — there is still a lot to play for in “instant” groceries. But if this latest round shows us anything, it’s that the most promising of these delivery companies will continue raising a lot more money to position themselves as consolidators within it.

Additional reporting: Natasha Lomas

Imagine a world where no one’s privacy is breached, no faces are scanned into a gargantuan database, and no privacy laws are broken. This is a world that is fast approaching. Could companies simply dump the need for real-world CCTV footage, and switch to synthetic humans, acting out potential scenarios a million times over? That’s the tantalizing prospect of a new UK startup that has attracted funding from an influential set of investors.

UK-based Mindtech Global has developed what it describes as an end-to-end synthetic data creation platform. In plain English, its system can imagine visual scenarios such as someone’s behavior inside a store, or crossing the street. This data is then used to train AI-based computer vision systems for customers such as big retailers, warehouse operators, healthcare, transportation systems and robotics. It literally trains a ‘synthetic’ CCTV camera inside a synthetic world.

It’s now closed a $3.25 million early-stage funding round led by UK regional backer NPIF – Mercia Equity Finance, with Deeptech Labs and In-Q-Tel.

That last investor is significant. In-Q-Tel invests in startups that support US intelligence capabilities and is based in Arlington, Virginia…

Mindtech’s Chameleon platform is designed to help computers understand and predict human interactions. As we all know, current approaches to training AI vision systems require companies to source data such as CCTV footage. The process is fraught with privacy issues, costly, and time-consuming. Mindtech says Chameleon solves that problem, as its customers quickly “build unlimited scenes and scenarios using photo-realistic smart 3D models”.

An added bonus is that these synthetic humans can be used to train AI vision systems to weed out human failings around diversity and bias.

Mindtech CEO Steve Harris

Steve Harris, CEO, Mindtech said: “Machine learning teams can spend up to 80% of their time sourcing, cleaning, and organizing training data. Our Chameleon platform solves the AI training challenge, freeing the industry to focus on higher-value tasks like AI network innovation. This round will enable us to accelerate our growth, enabling a new generation of AI solutions that better understand the way humans interact with each other and the world around them.”

So what can you do with it? Consider the following: A kid slips from its parent’s hand at the mall. The synthetic CCTV running inside Mindtech’s scenario is trained thousands of times over how to spot it in real-time and alert staff. Another: a delivery robot meets kids playing in a street and works out how to how to avoid them. Finally: a passenger on the platform is behaving erratically too close to the rails – the CCTV is trained to automatically spot them and send help.

Nat Puffer, Managing Director (London), In-Q-Tel commented: “Mindtech impressed us with the maturity of their Chameleon platform and their commercial traction with global customers. We’re excited by the many applications this platform has across diverse markets and its ability to remove a significant roadblock in the development of smarter, more intuitive AI systems.”

Miles Kirby, CEO, Deeptech Labs said: “As a catalyst for deeptech success, our investment, and accelerator program supports ambitious teams with novel solutions and the appetite to build world-changing companies. Mindtech’s highly-experienced team are on a mission to disrupt the way AI systems are trained, and we’re delighted to support their journey.”

There is of course potential for darker applications, such a spotting petty theft inside supermarkets, or perhaps ‘optimising’ hard-pressed warehouse workers in some dystopian fashion. However, in theory, Mindtech’s customers can use this platform to rid themselves of the biases of middle-managers, and better serve customers.